Measuring Entrepreneurial Mindset in Youth

MEASURING ENTREPRENEURIAL
MINDSET IN YOUTH:
LEARNINGS FROM NFTE’S ENTREPRENEURIAL MINDSET INDEX
November 2018
MEASURING ENTREPRENEURIAL
MINDSET IN YOUTH:
LEARNINGS FROM NFTE’S ENTREPRENEURIAL MINDSET INDEX
Thomas Gold and Sophia Rodriguez
November 2018
For more information about NFTE’s Entrepreneurial Mindset Index, please contact
Sophia Rodriguez, Director of Research & Analytics, at Sophia.Rodriguez@nfte.com
© 2018 NFTE (Network for Teaching Entrepreneurship). All rights reserved. You may make copies of and distribute
this work for noncommercial educational and scholarly purposes. For any other uses, including the making of
derivative works, permission must be obtained from NFTE, unless fair use exceptions to copyright law apply.
ACKNOWLEDGMENTS
This paper is the culmination of work spearheaded by the NFTE research team to identify and measure
entrepreneurial mindset. The authors owe an enormous debt to the individuals and organizations whose
support made this report possible. We would first like to acknowledge the generosity of EY, an organization
that has collaborated with NFTE for over 20 years and provided the funding and in-kind support needed to
create and test the Entrepreneurial Mindset Index (EMI). Additionally, we would like to recognize that support
from the Diana Davis Spencer Foundation and Citi Foundation has allowed NFTE’s focus on activating the
entrepreneurial mindset in youth to advance greatly over the last several years.
The NFTE Research Advisory Committee—Margaret Honey, Jill Kickul, Doug Ready, Sharon Weinberg and
Christoph Winkler -- provided early ideas around identifying and measuring entrepreneurial mindset in 2014
and 2015. Individual committee members continued to provide ongoing technical support and guidance during
the development of the tool in subsequent years.
Patrick Kyllonen, Kevin Petway and Meghan Brenneman from Educational Testing Service (ETS) provided
critical guidance, expertise, writing and analysis that led to the creation of the initial items for the EMI. Their
work, especially the analyses by Dr. Petway, set the ground for development of the final EMI and the analysis
in this paper.
A number of people reviewed and provided comments on different drafts of this paper, including Kåre Moberg,
Vishal Singh, Adam Ozimek, and Dante Deantonio. Their comments on an earlier draft helped us refine our
argument and focus in the paper. Chelsea Farley edited the final version of the paper.
Finally, we would like to thank all of the NFTE staff and teachers who helped with the development and testing
of the EMI. In particular, we’d like to acknowledge a few staff members who were instrumental to this paper.
Hannah Lieber reviewed drafts of the paper and provided comments on the psychometric aspects of the tool.
Zachary Tausanovitch conducted initial analyses of an earlier version of the EMI and helped craft the items
for this current version; Kim Smith, Jason Delgatto, Rob Pauls and Troy Pressens all helped in the process of
defining and crafting EMI domains. Shawn Osborne, President and CEO of NFTE, has been supportive of the
project since the early days. Barbara Clement and Eva Giorgi helped us take the paper to the finish line with
their design support.
212.232.3333 120 Wall Street, 18th Floor, New York, NY 10005 www.nfte.com
www.nfte.com
TABLE OF CONTENTS
EXECUTIVE SUMMARY................................................................................................1
INTRODUCTION.............................................................................................................3
PART 1—WHY ENTREPRENEURIAL MINDSET MATTERS........................................4
PART 2—THE ENTREPRENEURIAL MINDSET INDEX (EMI).....................................8
PART 3—EARLY EMI RESULTS:

EXPLORING CHANGES IN ENTREPRENEURIAL MINDSET....................................13
CONCLUSIONS............................................................................................................20
REFERENCES..............................................................................................................21
APPENDIX A.................................................................................................................23
APPENDIX B................................................................................................................29
APPENDIX C................................................................................................................35
APPENDIX D................................................................................................................38
ENDNOTES .................................................................................................................40
www.nfte.com
EXECUTIVE SUMMARY
Entrepreneurial thinking has become a mantra for many programs and initiatives that aim to prepare
young people for the future. This is because entrepreneurial skills—like the ability to take initiative and
think on your feet, critically solve problems, and communicate effectively—are increasingly viewed as
essential for career success.
But how do we know whether young people are acquiring these skills? Until now, there has been little
research to guide efforts to assess entrepreneurial thinking in youth. In recent years, the Network for
Teaching Entrepreneurship (NFTE) has been working to fill this gap by developing and testing the
Entrepreneurial Mindset Index (EMI)—a tool to measure attitudes, behaviors and beliefs associated
with being an entrepreneur. The EMI is currently in use with students in NFTE programs in schools
across the United States and abroad. Our new paper, Measuring Entrepreneurial Mindset, presents
findings on the utility, reliability and validity of the EMI, highlighting its potential as an innovative and
effective assessmenttool in the entrepreneurship education landscape.
Why Entrepreneurial Mindset Matters

The paper begins by outlining several reasons for the growing focus on entrepreneurial mindset,
including:
• A decline in entrepreneurship in the U.S. over the last several decades;
• Recognition that employers are actively seeking workers with entrepreneurial skills; and
• New research and policy initiatives underscoring the importance of noncognitive skills and
social and emotional learning.
Developing the EMI

The paper describes NFTE’s process for developing and piloting the EMI. It also presents the results
of robust statistical analyses examining the EMI’s reliability and validity. These analyses draw on
responses from about 4,000 students across the country who took the EMI twice during the
2016-2017 school year (once near the beginning of NFTE’s Owning Your Future entrepreneurship
course and once near the end).
Overall, based on a variety of statistical tests, the findings indicate that the EMI is a reliable
and valid assessment tool to measure entrepreneurial mindset. The version of the EMI that
we tested included six constructs: future orientation, opportunity recognition, communication and
collaboration, critical thinking and problem solving, comfort with risk, and creativity. All of these had
high levels of reliability and internal consistency. We also found the six constructs to be collectively
reliable and consistent—suggesting they are effectively measuring the broader concept of
entrepreneurial mindset.
Early EMI Results: Exploring Changes in Entrepreneurial Mindset

The final section of the paper explores how students’ entrepreneurial mindset changed during their
participation in the NFTE course.* This analysis showed little difference between average pre and
post scores, both overall and for each individual construct. But the averages masked a fair amount of
variation in students’ EMI scores: We found that roughly half of the students in the study experienced
212.232.3333
EMI 120 Wall
growth, while the other half Street, 18athdecline
experienced Floor, New York,
in their NY 10005
entrepreneurial mindset. www.nfte.com
1 www.nfte.com
This variation allowed us to examine how changes in the EMI relate to growth in other, relevant
entrepreneurial attitudes and behaviors. For this analysis, we relied on measures of entrepreneurial
self-efficacy, entrepreneurship as an applied skill, and entrepreneurial intentions. We found that,
compared to students who experienced a decline in entrepreneurial mindset, those whose
entrepreneurial mindset improved:
• Are twice as likely to have growth in their entrepreneurial self-efficacy (i.e., 2.0 odds ratio);
• Are twice as likely to think about entrepreneurship as a skill that can be applied in any career
(i.e., 2.0 odds ratio); and
• Have a 70 percent greater chance of wanting to own a business at the end of the course
(i.e., 1.7 odds ratio).
We also wanted to understand whether the EMI was related to entrepreneurial knowledge. We found:
• The greatest gains in mindset happened in classrooms with the highest growth in
entrepreneurial learning, while the lowest mindset growth occurred in the
lower-performing classrooms.
Taken together, these findings indicate that EMI scores are meaningful and related to other
areas associated with entrepreneurial thinking and behavior. This information is important for
establishing the value and utility of the EMI. It also provides suggestive evidence that growing
entrepreneurial mindset may have a positive effect in getting youth to see entrepreneurship and
self-employment as a possible career path and something that can be learned and developed.
Similarly, the results suggest that helping youth develop the entrepreneurial mindset may in fact
increase their intentions to start a business.
More research is needed to further explore the conditions that promote EMI growth. We hope the
current findings serve as a starting point for productive dialogue and additional exploration, as
entrepreneurship education continues to flourish.
Entrepreneurial Mindset Index Domains

This paper explores the first six domains below, consisting of Likert Agreement Scale items.
The other two domains in grey below consist of Situational Judgement Tests.
*Our dataset for these analyses included 3,661 students who took both the pre-and post-EMI in the 2016-2017 school year and who
answered enough questions to generate an overall EMI mean score in both time periods.
2 www.nfte.com
INTRODUCTION
Increasingly, the global economy demands that young people enter the workforce not only with a college
degree, but also with a set of transferrable, entrepreneurial skills and attitudes that can help them succeed in
almost any job or industry. This includes the ability to take initiative and think on your feet, to critically solve
problems, and to communicate effectively. Learning these and other skills that are part of the entrepreneurial
mindset is central to becoming career-ready. As a result, entrepreneurial thinking has become the mantra for
many organizations interested in preparing youth for the future.
But how do we assess whether young people are acquiring these crucial skills? Until now, there has been little
research to guide efforts to accurately measure entrepreneurial mindset. During the last several years, the
Network for Teaching Entrepreneurship (NFTE) has been working to fill this gap by developing and testing the
Entrepreneurial Mindset Index (EMI)—a tool to measure the attitudes, behaviors and beliefs associated with
being an entrepreneur. The EMI is currently in use with students in the NFTE entrepreneurship program in
schools across the United States and abroad. This paper presents new findings about the utility, reliability and
validity of the EMI, highlighting its potential as an innovative, effective assessment tool in the entrepreneurship
education landscape.
The paper is divided into three sections: (1) First, we set the stage by discussing why teaching and measuring
entrepreneurial mindset matters. This section includes a description of NFTE’s theory of change about how
to activate the entrepreneurial mindset in youth. (2) Second, we describe the development of the EMI. This
section presents the results of a robust statistical analysis examining the instrument’s reliability and validity.
(3) Finally, we explore how changes in students’ entrepreneurial mindset, as measured by the EMI, appear to
correlate with related measures of entrepreneurial attitudes, intentions and knowledge. This provides important
evidence about the degree to which the EMI is indeed measuring something meaningful and relevant to
entrepreneurship education.
3 www.nfte.com
PART 1—WHY ENTREPRENEURIAL MINDSET MATTERS
As we describe below, the growing focus on teaching entrepreneurial skills stems from both clear needs in
the economy and an expanded understanding of the skills young people need to succeed in school, work
and beyond.
Nurturing Entrepreneurship
In recent decades, entrepreneurship – business startup – has been on the decline in the United States.1 In the
late 1980s, for example, new firms accounted for 13 percent of all businesses. Twenty years later, that number
had dropped to 8 percent. The percentage of adults owning their own businesses has also dropped since the
1990s.2 The decline in entrepreneurship is concerning not only because of its potential negative impact on
new job growth but also because of its cultural significance. America was founded on the ideals of liberalism
and the free market. At its best, America has welcomed people from all over the world and given them the
opportunity to build new businesses from scratch and make big ideas a reality. The loss of entrepreneurship
is the loss of something quintessentially American.
Building Skills that Employers Need

Despite the decline in actual startup activity, the skills associated with being an entrepreneur are increasingly
critical to thrive in the 21st century economy.3 Employers are actively looking for entrepreneurial workers—with
skills like creativity, critical thinking and complex problem solving.4 In fact, one in three employers polled in
the United States said that they prefer hiring candidates with at least some entrepreneurial experience.5 As a
result, teaching entrepreneurial thinking should be seen as critical to strengthening human capital and making
individuals more employable and competitive.6
Teaching Key Noncognitive Skills that Help Students Succeed

The focus on entrepreneurial mindset is consistent with a variety of other efforts in K-12 education to teach
noncognitive skills and promote social and emotional learning (SEL).7 As the authors of a 2015 Economic
Policy Institute report explained,
“Broadly, these skills encompass those traits that are not directly represented by cognitive skills or by
formal conceptual understanding, but instead by socio-emotional or behavioral characteristics that are not
fixed traits of the personality, and that are linked to the educational process, either by being nurtured in
the school years or by contributing to the development of cognitive skills in those years (or both).” 8
There is growing evidence that such skills can be critical for academic and career success. Various studies
have demonstrated the impact of addressing noncognitive skills in the classroom. According to one widely
cited meta-analysis, programs that focus on non-academic skills not only improved attitudes and behaviors,
but also academic performance. The authors found that students participating in SEL programs showed an
11-percentile gain in academic performance compared with students in control groups.9
In response to this research, many states and school districts have begun to explicitly target noncognitive
skills as part of their K-12 curricula. This trend was accelerated by the passage, in 2015, of the Every Student
Succeeds Act, which provided states with more flexibility in how they understand student success and how
they measure school quality. Similarly, a growing number of foundation initiatives and out-of-school programs
are focused on building non-cognitive skills, including many that are part of the entrepreneurial mindset.
4 www.nfte.com
Advances in Entrepreneurship Education
For all of these reasons, entrepreneurship education has expanded in the United States and globally. Much of
this growth has taken place at the college and university level, where the number of entrepreneurship majors
and clubs has grown considerably over the last decade.10 That said, there have been promising developments
in entrepreneurship education at the K-12 level as well. For example, 84 percent of states currently report
having K-12 standards, guidelines or proficiencies in entrepreneurial education, up from only 38 percent in
2009. Likewise, 36 percent of states require that high schools make entrepreneurial coursework available,
up from 10 percent in 2009.11
We see similar trends outside of the United States. In 2016, the European Union unveiled the EntreComp
framework, building off of the organization’s identification of entrepreneurship as one of eight lifelong learning
competencies—and one that is necessary for a knowledge-based society. The initiative, which promotes
entrepreneurship education in schools, aims to “raise consensus among all stakeholders and to establish
a bridge between the worlds of education and work.”12
The Next Challenge: Effectively Measuring Entrepreneurial Mindset in Youth

As programs and courses in entrepreneurship proliferate, there are pressing questions about how to measure
their outcomes. The goals of entrepreneurship education are not always the same for every program. Some
aim to help participants create actual startups13, while others are primarily geared toward the development of
core skills associated with being an entrepreneur. These skills are relevant in a wide range of settings, across
industries, including corporate intrapreneurship—that is, entrepreneurial activity that leads to innovation within
large established organizations.14 While programs vary, understanding and measuring how they influence the
way students think—their mindset—is critically important for gauging success.
Researchers are just beginning to tackle the challenges of measuring entrepreneurial mindset and other
noncognitive skills. These skills are difficult to assess because they do not represent a clear transfer of
knowledge. Unlike with traditional academic areas like math or reading, measuring entrepreneurial mindset
requires one to understand how a program influenced a student’s “patterns of thought, feelings and behavior”
and not just facts that they learned in a class.15 Innovative assessment tools are needed.
A number of efforts have been made to develop assessments of entrepreneurial mindset.16 One of the
clear challenges that these assessments face is a lack of agreement about the definition of entrepreneurial
mindset.17 What is most relevant about how an entrepreneur thinks? While some identify the importance of
comfort with ambiguity,18 others are more focused on perseverance19 or pro-activeness.20 Many assessments
look at a mix of both soft and hard skills, including information related to owning and running a business.
Most of these existing assessments are geared toward adults, including those that are already engaged
in business and new ventures. Others are built to assess university students, including many focused on
specific academic majors, like engineering or business. Few of these assessments are appropriate for youth
in middle and high school because they focus largely on intent to start a business—whereas entrepreneurship
education at the middle and high school level is mainly geared toward activating an interest in
entrepreneurship and entrepreneurial thinking. One of the best efforts to develop an assessment of
entrepreneurship education for school-aged youth has been the ASTEE project in Europe (ASTEE 2014).21
NFTE’s effort to develop an assessment of entrepreneurial mindset draws from many of the lessons learned
from the ASTEE study. As outlined below, our work also relies on careful consideration of the short- and
long-term student outcomes that NFTE’s programs are designed to influence.
5 www.nfte.com
NFTE’S Theory of Change
NFTE is an internationally recognized nonprofit that strives to activate the entrepreneurial mindset in young
people through programs in middle and high schools. Each year, NFTE serves over 20,000 students in the
United States and another 40,000 abroad. NFTE’s signature program, Owning Your Future, is a semester- or
year-long course that teaches students the fundamentals of business startup through classwork, experiential
exercises, engaging with volunteers and the development of a business plan.22 At the end of the course,
students are required to present their business plans to their class. Many NFTE students continue their
interest in entrepreneurship and compete for prizes in business plan competitions sponsored by NFTE.
Figure 1 on the next page maps out the NFTE theory of change. As shown in the first part of the figure,
over the course of our programs, students learn the fundamentals of business startup and entrepreneurship
through classroom instruction, with a strong focus on experiential activities. This includes games, short
projects and lessons that give students hands-on experiences and teach them concepts like recognizing
opportunity, generating a profit, and marketing. The students also gain real-world experience by working
closely with volunteers and mentors who are often business leaders from the community. These interactions
help the students take the abstract concepts they are learning in the classroom and see how they function in
the real world.
As shown in the second part of the figure, these program activities are expected to help students develop their
entrepreneurial mindset and begin thinking like an entrepreneur. This becomes stronger toward the end of
program as they complete their business plan and present it to their class and, for many students, to external
juries at regional and national business plan competitions.
The third part of the figure shows that when students have developed the entrepreneurial mindset, it leads
to other changes in attitudes and behavior, including promoting a more positive disposition toward business
startup and entrepreneurship. Previous studies have demonstrated the connection between entrepreneurship
education and intentions to start a business.23 We also expect that students begin to view entrepreneurship as
a possible career choice.24
Finally, as shown in the last section of the figure, changes in students’ attitudes and behaviors are expected to
support their ability to start a business or otherwise demonstrate career success. We have strong suggestive
evidence from previous research that students who have participated in the NFTE program are more likely
than others to become business owners: A 2017 study found that 25 percent of NFTE alumni reported starting
their own business. This is over ten times the national average for youth businesses in the United States and
more than twice the overall national average.25
Utilizing data from NFTE’s core program, Owning Your Future, this paper focuses on the intermediary steps
in our theory of change: the relationship between entrepreneurial mindset activation and entrepreneurial
attitudes and behaviors. Before we present those findings, however, we describe the development of our
Entrepreneurial Mindset Index as well as the results of our analysis of its reliability and validity.
6 www.nfte.com
Figure 1. NFTE’s Theory of Change: How Teaching the Entrepreneurial Mindset Supports
Student Success
Focus of this Paper
Entrepreneurial Entrepreneurial Entrepreneurial Entrepreneurial

Education Mindset Attitudes and Actualization and
Activation Behaviors Career Success
• Knowledge of
business startup • Positive growth • Perspective of • Actualization of
fundamentals in mindset at the entrepreneurship business start-up
• Experiential/ end of NFTE as as a possible • Employment rate
project-based measured by career
NFTE’s EMI • Career success
learning • Perspective of
instrument entrepreneurship • Career satisfaction
• Business
ideation as a skill that can
be applied to
• Creation of
other careers
business plan
(individually/ • Entrepreneurial
in a team) intent
• Development and • Career readiness
presentation of
business pitch
• Feedback and
support from
business
professional
7 www.nfte.com
PART 2—THE ENTREPRENEURIAL MINDSET INDEX (EMI)
Deciding What to Measure

In 2014, NFTE started in earnest the process of developing an assessment of entrepreneurial mindset for
youth. Leading up to this moment, NFTE had identified the development of entrepreneurial mindset as a major
goal of the organization, in part from a realization of the importance of developing these skills to prepare
students for the 21st century workforce. In addition, earlier studies of NFTE had highlighted the program’s
ability to affect noncognitive skills, such as students’ internal locus of control.26 This evidence helped drive a
focus on entrepreneurial mindset as a primary programmatic outcome.
As a first step in the development of the EMI, we had to define the concept we were trying to measure. What is
entrepreneurial mindset? How do we know when a student has “learned” it? Can it, in fact, be learned? During
the first phase, we identified eight key skills, attitudes and behaviors that the EMI should assess:
• future orientation
• comfort with risk
• opportunity recognition
• flexibility and adaptability
• initiative and self-reliance
• creativity and innovation
• critical thinking and problem solving
• communication and collaboration
To identify these key domains—that is, the constructs we sought to measure—we reviewed the literature
on entrepreneurship education, 21st century skills, noncognitive factors, social and emotional learning, and
character education. We also conducted interviews and focus groups with staff, teachers, students, alumni
and other stakeholders to understand outcomes that were most closely associated with the NFTE program.
In addition, we looked for opportunities to align potential EMI domains with existing competencies identified
by the Collaborative for Academic, Social, and Emotional Learning (CASEL), the National Research Council
(NRC), and the Cognitive Ability or Big 5 Personality traits.
During this early period, we decided that the assessment should focus more on softer skills, like
communication and problem solving, which led us to eliminate some domains that had been selected earlier,
like financial and digital literacy. While we saw these skills as important to entrepreneurship and business
startup, we wanted the assessment to address essential cross-cutting noncognitive skills as much as possible.
In June 2014, we convened a working group of top researchers in the fields of entrepreneurship, psychology,
education policy, and education technology to discuss the development of an assessment of entrepreneurial
mindset. Key findings from the meeting included an affirmation of the importance of developing
entrepreneurial mindset in youth as a core programmatic goal for NFTE. In addition, there was a clear
agreement from the group that in order to measure entrepreneurial mindset, we needed to identify
innovative assessment approaches. Finally, attendees felt that the information from the assessment should be
used not just for the purposes of program evaluation, but also to provide formative feedback that could inform
and strengthen instruction.
8 www.nfte.com
In response to the working group’s focus on innovative assessment approaches, we initiated a collaboration
with the Educational Testing Service (ETS) in late 2014. This collaboration allowed us to finalize the
assessment constructs, develop survey items (i.e., questions) to measure them, and begin to analyze results
from piloting the assessment. The ETS team brought expertise not only in psychometrics and education
research, but also in assessments of noncognitive factors and workforce development skills. The same team
from ETS had worked on a similar project to develop another noncognitive assessment—the Mission Skills
Assessment (MSA)—that is now in wide use in private, independent schools.27
Developing and Piloting Items

After finalizing the assessment constructs, we worked with ETS to draft assessment items during the spring
and summer of 2015. During this initial phase, we developed over 80 items, including mostly Likert type
(agreement scale) and biographical questions (“I have presented an idea in front of my class during the
last month”).28
As part of the development of Likert items, the ETS team conducted “cognitive labs” with NFTE students
during our summer Bizcamps in New York City in 2015. A cognitive lab is a method to study the mental
processes one uses when completing a task, which in this case was answering draft assessment items. During
the labs—which involved 11 NYC high school students from low-income backgrounds—participants were
asked to answer questions and then discuss their responses. This process allowed us to change language in
some of the items that students found confusing or unclear.29 It also helped us address other issues like poor
choice of words, culturally specific phrases and bad grammar, all of which can influence item responses.
We finalized the items and then piloted the EMI during the 2015-2016 school year with around 1,000 students
participating in NFTE’s Owning Your Future course in about 15 NYC public high schools. In total, 73 Likert
items were piloted with students in the fall and spring of the school year. In addition, 14 Situational Judgement
Tests were piloted during this phase.30
The assessment was administered online. Student responses were secure and largely anonymous. (They
provided a self-created key using information like their birthdate, first few letters of their mother’s maiden
name, first characters of street address, etc.). This latter step was done to match pre- and post-tests,
where needed.
This initial pilot was conducted mainly to begin to assess the reliability of the different constructs and to
reduce the number of items. Based on the initial round of testing, we were able to cut the number of Likert
items from 73 to 30, as many questions proved to be redundant. In addition, we were able to determine that
the six constructs that used Likert items were internally consistent, meaning that individual students on the
whole responded similarly to items that fall within the same domain. We selected these constructs—future
orientation, opportunity recognition, communication and collaboration, critical thinking and problem solving,
comfort with risk, and creativity—to roll out in the fall of 2016.31
To address the remaining two constructs (flexibility and adaptability, initiative and self-reliance), we had used
the more experimental Situational Judgement Tests. However, these items did not hold up in our first round
of testing. Subsequently, we rewrote these items in collaboration with teachers for use in the 2017-2018 school
year. Those results are still in the process of being analyzed for a separate study. The rest of this paper
addresses the results of the six reliable constructs that were piloted in all NFTE schools during the
2016-2017 school year.
9 www.nfte.com
Assessing the EMI’s Reliability and Validity
This paper uses results from about 4,000 students with matched data (i.e., students who both took our
pre- and post-EMI). The EMI was administered to all participants in NFTE’s Owning Your Future
entrepreneurship course in 2016-2017. Similar to the pilot test the previous year, the EMI was administered
near the beginning and end of the course—either in the fall and spring (for students taking the year-long
course) or early and late in a single semester (for those taking the one-semester course).
As noted above, this paper focuses on the six constructs that had been previously identified as reliable with
items that demonstrate internal consistency. Figure 2 on the next page lists the definition and number of items
for each of these six constructs, along with corresponding sample items from the EMI.
Similar to other efforts in entrepreneurship education to measure mindset (e.g., ASTEE 2014), we employed
structural equation modeling (SEM) as our umbrella statistical technique. We use SEM to determine how the
items are related to the six constructs32 we defined to make up the entrepreneurial mindset. SEM is a popular
method for behavioral researchers to measure what are called latent constructs or factors that are theoretical
in nature but not directly observable. SEM helps us understand whether a set of observed variables—that is,
the questions or items that we ask the respondents—are correlated to the theoretical constructs that we are
interested in measuring.33 Utilizing a variety of techniques, from confirmatory factor analysis (CFA) models to
regression models, SEM enables us to assess whether questions within each construct are reflective of their
respective EMI domain and whether they can hold together statistically as an “overall model” to measure
entrepreneurial mindset.
Figure 2.
EMI Domains
and Sample Items
10 www.nfte.com
In this paper, we focus primarily on using CFA models to test the following assumptions:
• The six individual constructs as well as the overall EMI model hold together statistically for the pre-EMI
and post-EMI (model fit).
• The six constructs and the EMI are consistently measuring similar constructs over time (reliability).
• The questions within the six constructs are reflective of that construct overall, and collectively, are
statistically related to an overall theoretical model of entrepreneurial mindset (internal consistency
and convergent validity).
• The six constructs are not reflecting other constructs not included in the model (discriminant validity).
• Different groups of students approach the questions similarly (measurement invariance).
Below, we provide the results of our analyses, which include the relevant estimates alongside the
recommended ranges of the statistical tests. Overall, the findings establish the EMI as a consistent and
valid assessment tool to measure entrepreneurial mindset.
All six of the constructs tested had high levels of reliability and internal consistency based on relevant tests.
In particular, they all exceeded the recommended Cronbach’s alpha and composite reliability cut-off of 0.70.
We also found the six constructs to be collectively reliable and consistent, which was important for our later
analyses that examined overall EMI scores. We were unfortunately unable to establish discriminant validity
for all of the constructs. Particularly, there seemed to be a high correlation between the Creativity and
Innovation and Critical Thinking and Problem Solving constructs, as well as the Creativity and Innovation
and Opportunity Recognition constructs. This is most likely the result of actual similarities between these three
constructs and what they are attempting to measure. In the end, the overall EMI model emerged with a good
statistical fit based on a number of indices. See Table 1 for an overview of the analyses of our pre-EMI and
Appendix A for details.
Model Fit, Reliability, Internal Consistency, Convergent and Discriminant Validity

Table 1 provides a detailed description of our model fit indices as well as estimates that detail reliability,
internal consistency, and convergent and divergent validity. (Table 1 presents findings from the post-test EMI.
We conducted this analysis for both the pre- and post-test. Findings from the pre-test are similar and presented
in the Appendix A.)
11 www.nfte.com
Table 1. Reliability, Internal Consistency, and Model Fit of Post-EMI (Six Constructs)
Measurement Invariance
We conducted measurement invariance analyses to understand how different subgroups of students, in terms
of ethnicity, gender, grade level and course requirement, approached the EMI questions. These tests help us
determine if there is any bias in the way we have framed the questions that would impact the results. Our
analysis found that students in the sub-groups interpreted the questions for the six constructs similarly in both
the pre- and post-EMI tests. This was the case even in the most rigorous invariance testing that we had
applied. See Appendix A for more
212.232.3333 120details.
Wall Street, 18th Floor, New York, NY 10005 www.nfte.com
12 www.nfte.com
PART 3—EARLY EMI RESULTS: EXPLORING CHANGES IN
ENTREPRENEURIAL MINDSET
The previous section detailed the statistical testing of the EMI, including tests examining the instrument’s
reliability, validity and measurement invariance. Overall, these analyses confirm that the EMI is measuring the
six constructs consistently and accurately. The final section of this paper begins exploring results from the EMI
in 2016-2017, looking at how students’ entrepreneurial mindset changed between the pre- and post-tests. This
allows us to then examine how changes on the EMI relate to growth in other, relevant entrepreneurial attitudes
and behaviors.
This analysis is key to establishing “nomological” validity, an approach that is used to help establish the
meaningfulness of unobservable constructs by linking them with measures that share a theoretical connection.
In this case, we hypothesized that the greatest growth in entrepreneurial mindset would be in classrooms with
the highest levels of entrepreneurship knowledge learning. While these analyses are exploratory in nature, we
believe they are an important way to establish the value and utility of this innovative assessment.
About the Sample

Our dataset for these analyses encompasses 3,661 students in NFTE’s signature entrepreneurship course
who took both the pre- and post-EMI in the 2016 – 2017 school year and who answered enough questions to
generate an overall EMI mean score in both time periods.34 In addition to the EMI, we also collected a host of
data that we used for our analyses. Table 2 on the next page describes the sample, including demographics,
level in school, home state and other characteristics.
We utilize demographic and course variables as controls in our regression analysis. We also incorporate
the fixed effects of the different states for each of the NFTE programs. As you see in the table, the EMI was
administered in NFTE programs in 14 different states, including the District of Columbia. We include these
fixed effects in the model to control for heterogeneity in education policies—which can vary considerably from
state to state, potentially impacting the way the program is implemented. When we analyzed changes in the
entrepreneurial mindset among the students in our sample, we found little difference between average pre and
post scores—both overall and within each of the domains (see Figure 3 on page 14).
But these averages masked a fair amount of variation in students’ scores. As shown in Figure 4 on page 15,
roughly half of the students indicated a positive mindset change, while the other half indicated a negative
mindset change. There are several potential factors that account for this even split in mindset change, the
most logical being variation in the implementation of the NFTE program. While this paper does not go into
the factors behind the variation, the relatively even distribution of students with mindset growth and decline
presented us with an opportunity: Specifically, it allowed us to use these two categories of students as
subgroups and compare whether they varied in terms of their entrepreneurial attitudes, behaviors
and knowledge.
Thus, our approach in the following section is to examine the correlation between mindset growth and other
areas of change that would be consistent with NFTE’s theory of change. Our study does not utilize a
comparison or control group of students who did not participate in the NFTE program. Rather, we focus on
the differences between students who demonstrate mindset growth compared to those who show
mindset declines.
13 www.nfte.com
Table 2. Descriptive Statistics for the Sample
Figure 3. Entrepreneurial Mindset Index Results, 2016-2017
14 www.nfte.com
Figure 4. Distribution of
Entrepreneurial Mindset
Change between
Pre and Post
Analytic Approach
For this section, we used a variety of regression analyses—Ordinary Least Squares (OLS), Logistic, and
Analysis of Variance (ANOVA), depending on whether our dependent measure was ordinal, binary, or
categorical, respectively. Our primary model compared those who grew their mindset with those who did not,
while controlling for a range of other factors, including their EMI score at the beginning of the course, their
demographics, school and classroom. This model treated our EMI change measure as a binary independent
variable for our OLS and Logistic regressions (see Model 1 below). Students were coded as “1” if they had
some level of positive change in their EMI mean score between the beginning and end of the NFTE course,
and “0” if they had some level of negative change; there were a negligible number of students with static
scores, which were omitted. We also wanted to explore whether EMI changes were associated with changes
in entrepreneurial knowledge. To do so, we used ANOVA analyses, in which students’ EMI mean change
scores were analyzed as a dependent variable, with the primary independent variable being whether they
were in classes displaying low, moderate or high levels of learning in entrepreneurial knowledge (see Model 2).
Model 1.
Yis = β1(EM Change Binary)is + β2(Pre)is + β3(Dems)is + β4(Course)is + ai + uis
Model 2.
Y (EM Change Mean)is = β1(Group)is+ β2(Pre)is + β3(Dems)is + β4(Course)is + ai + uis
In both models:
• Pre represents the mean EMI score of the student at the beginning of the course to account for
student-level variation in their entrepreneurial mindset at the outset. Additionally, we included a
student’s response to the entrepreneurial and career attitude/behaviors at the beginning of the
NFTE course in our inference analyses to establish a baseline model.
15 www.nfte.com
• Dems represents student-level characteristics, including ethnicity, gender, whether the student is an
immigrant, and whether the student has a close family member who is an entrepreneur.
• Course represents course-level characteristics, including whether the NFTE course was a full year or
semester, required or elective, standalone or embedded into another course (such as Business), and
whether the course was implemented at a high school or middle school.
• ai is the state-invariant fixed effects to account for any unobserved variation in NFTE courses due to
heterogeneity in state education policies.
• uis is the error term
Previous entrepreneurship education studies have used similar demographic and course controls, establishing
them as important covariates to include in such analyses (e.g., Moberg 2012, Hayes and Richmond 2017).
We conducted goodness of fit tests as well as checked for multicollinearity in all of our models. We also
conducted additional sub-group analyses that examined students who changed their attitudes and behaviors
in these different categories, from negative to positive during the NFTE course, compared to students who
went from positive to negative or students who remained static in their negative attitudes and behaviors.35
How Do EMI Scores Relate to Other Relevant Entrepreneurial Attitudes

and Behaviors?
This section explores how EMI scores are related to other factors that are relevant and important to
entrepreneurship and entrepreneurial thinking. For this analysis we relied on established measures of
entrepreneurial self-efficacy, entrepreneurship as an applied skill, and entrepreneurial intentions. All of these
measures have been used in other studies of entrepreneurial thinking.36
Entrepreneurial Self-Efficacy (ESE) refers to an individual’s belief that they have the knowledge and skills
to start their own business.37 Our measure of ESE was captured by a separate Likert agreement scale item
(“I feel I have the skills and knowledge required to start a new business”). We focused on ESE growth over
the course of the NFTE experience, as we were most interested in understanding whether growth in ESE was
related to growth in entrepreneurial mindset.
The focus on self-efficacy in entrepreneurship research is based on the work of Bandura, whose social
cognitive studies looked at how individuals assessed their own “competencies and ability to overcome adverse
conditions and obstacles and the belief that future actions will be successful”.38 These beliefs in turn affect
“what challenges to undertake, how much effort to expend on the endeavor (and) how long to persevere in the
face of difficulties.”39
Entrepreneurship as an Applied Skill measures students’ self-report of whether they feel that
entrepreneurship is a skill that can be used in any career.40 We asked this item as part of a broader question
about students’ views of entrepreneurship. When assessing this relationship, we looked at both how students
answered the item at the end of their NFTE experience, as well as the sub-group of students who changed
from not feeling that entrepreneurship was an applied skill at the beginning of their course to feeling that it was
an applied skill at the end of it.
Entrepreneurial Intention measures students’ self-report of whether they would like to own their own
business.41 When assessing this relationship, we looked at both how students answered the item at the end
of their NFTE experience, as well as the sub-group who changed from not wanting to own a business at the
beginning of their course to wanting to own a business at the end of it.
16 www.nfte.com
We found that, compared to students who experienced a decline in entrepreneurial mindset, those whose
entrepreneurial mindset improved:
• Are twice as likely to have positive entrepreneurial self-efficacy (ESE) growth

• Are twice as likely to think about entrepreneurship as a skill that can be applied in any career
• Have a 70 percent greater chance of wanting to own a business at the end of the course
EMI and Entrepreneurial Self-Efficacy

We used ESE to help validate the growth measure of the EMI. This relationship is important to establishing
predictive validity of the EMI. Given the prevalence of ESE in entrepreneurship and entrepreneurship
education studies, it is important to know whether students’ growth in mindset is closely related to their growth
in self-efficacy. We theorize that students’ confidence in their entrepreneurial ability is tied to their exposure to
entrepreneurship due to the experiential nature of NFTE, which in turn is related their entrepreneurial mindset.
Our theory is that this belief in one’s own capacity is based on a self-assessment of their past experiences and
knowledge and is, in the end, a measure of their readiness to perform that task.
In our analyses, we found a significant relationship between growth in entrepreneurial mindset and growth in
ESE (see Appendix B). Students who experienced a positive entrepreneurship change between the beginning
and end of NFTE had almost a half-point increase in their ESE six-point scaled measure, controlling for our
pre-EMI measure, demographic and course covariates, as well as incorporating state fixed effects. This
association was statistically significant, supporting our theory that EMI growth and ESE growth are related.
When analyzing ESE as a binary construct (1 = positive growth between beginning and end of NFTE,
0 = negative growth), we found that students with a positive entrepreneurial mindset had twice the odds
of having positive ESE growth, which was a statistically significant relationship.42
EMI and Entrepreneurship Attitudes and Intentions

Students who show positive growth in their entrepreneurial mindset have twice the odds of thinking about
entrepreneurship as a skill that can be used in any career at the end of NFTE, controlling for their views at
the beginning of the course, their mindset at the beginning of the course, as well demographic and course
controls and fixed effects (see Appendix B). This relationship between positive mindset growth and positive
entrepreneurial attitude is statistically significant.
Additionally, positive growth in entrepreneurial mindset is also a predictor for students who experience a
change in their attitude toward entrepreneurship as an applied skill. Compared to students who demonstrate
a negative mindset change, students who grow their mindset have statistically significantly greater odds of
changing their belief that entrepreneurship is a skill that can’t be used in any career to believing that it can.
(see Appendix B).
Finally, changes in entrepreneurial mindset may also lead to students having the intention to start a business43.
Students who show positive growth in their entrepreneurial mindset multiply their odds of wanting to own a
business by 1.7 at the end of NFTE, controlling for their views at the beginning of the course, and including
relevant covariates and fixed effects (see Appendix C).44 We believe that the relationship between EMI change
and entrepreneurship attitudes and intentions are critical in helping validate the EMI as a tool that is relevant to
the entrepreneurship education landscape.
17 www.nfte.com
How Do EMI Scores Relate to Entrepreneurship Knowledge?
While the above analyses suggest that there is a correlation between entrepreneurial mindset and
entrepreneurial behaviors and attitudes, we also wanted to understand whether the EMI was related to other
measures of entrepreneurship and, in particular, entrepreneurial knowledge. Entrepreneurial knowledge
was based on a 27-item cognitive assessment administered to NFTE students. Because the data lacked an
identifying variable that would have connected the EMI and entrepreneurial knowledge scores, we used
classroom mean knowledge scores instead. We were able to collect both the change in knowledge scores
between the beginning and end of NFTE, as well as the knowledge scores at the end of the course. We
calculated z scores with the classroom-level mean scores to create four classroom-level ranks, where 1
signified classrooms with the highest levels of entrepreneurship learning and 4 signified the lowest levels of
entrepreneurship learning.
We found:
• The greatest gains in mindset happened in classrooms with the highest growth in
entrepreneurial learning, while the lowest mindset growth occurred in the l
ower-performing classrooms.
Our theory was that students in classrooms with higher average gains in entrepreneurial knowledge
would experience a commensurate increase in their entrepreneurial mindset. This is not because
entrepreneurial learning is necessarily an antecedent to entrepreneurial mindset. Rather, our theory rests
on the assumption that students in these higher-performing classrooms were in fact experiencing better, more
effective instruction, most likely by a more effective teacher, and therefore receiving a stronger “dosage” of
entrepreneurship training.
In keeping with this theory, we found that the highest mindset growth occurred in classrooms with the highest
level of entrepreneurship knowledge learning. Likewise, we found the lowest mindset growth—in fact, declines
in mindset—in classrooms with the lowest performance on the entrepreneurship knowledge assessment. The
difference between these EMI changes was statistically significant and included all our controls and state fixed
effects (see Appendix D). Figure 5 presents these results.
Figure 5. ANOVA Fitted Values Model of EMI Change Scores by Classroom Entrepreneurship
Knowledge Assessment Levels (Lowest Level = 1, Highest Level = 4)
18 www.nfte.com
The analysis yielded identical results when we looked at the classrooms’ average change in knowledge
scores between the beginning and end of NFTE. The summative scores at the end of the year helped establish
classes that had grasped the most entrepreneurship knowledge. The change scores helped account for the
variation in student and teacher ability and classroom contexts.
These results further help us establish the validity of the EMI assessment. They demonstrate that the EMI
scores increase in contexts where we would expect it to—in this case, when students are in a high-quality
classroom setting.
19 www.nfte.com
CONCLUSIONS
The results of this study help establish the EMI as a valid and reliable instrument to measure entrepreneurial
mindset. The findings show not only that the EMI can reliably measure six different aspects of
entrepreneurial mindset, but that this measure is meaningful and related to other areas associated with
entrepreneurial thinking and behavior, particularly entrepreneurial self-efficacy and entrepreneurial knowledge.
These findings provide suggestive evidence that growing entrepreneurial mindset may have a positive effect
in getting youth to see entrepreneurship and self-employment as a possible career path and something that
can be learned and developed. Similarly, the results suggest that helping youth develop the entrepreneurial
mindset may in fact increase their intentions to start a business.
While the intentions do not necessarily translate directly into action, they provide a potential for more
entrepreneurship and entrepreneurial behavior. And NFTE’s own experience shows that providing
entrepreneurship education to middle and high school students has the potential to lead to more
enterprising behavior.
More research is needed to further explore the conditions that promote EMI growth. We are particularly
interested in examining the impact of NFTE’s program on students’ mindsets and other attitudes, beliefs and
behaviors, using a rigorous experimental or quasi-experimental approach that employs a suitable comparison
or control group. We also want to design studies that explore the other aspects of NFTE’s theory of change
around the entrepreneurial mindset, especially those that help us better understand the antecedents to
mindset growth.
As entrepreneurship education flourishes in schools and universities, additional research and evaluation have
a crucial role to play in developing effective programs. We hope these findings serve as a starting point for
productive dialogue and further exploration.
20 www.nfte.com
REFERENCES
Akmaliah, Zaidatol & Lope Pihie, Zaidatol. (2018). Entrepreneurship as a Career Choice: An Analysis of Entre-
preneurial Self-Efficacy and Intention of University Students.
http://publications.jrc.ec.europa.eu/repository/bitstream/JRC101581/lfna27939enn.pdf
Bacigalupo, M., Kempylis, P., Punie, Yves, & Van den Brande, G. (2016). EntreComp: The Entrepreneurship
Competence Framework. JCR Science for Policy Report. European Commission.
Bae, T. J., Qian, S., Miao, C., & Fiet, J. O. (2014). The relationship between entrepreneurship education and
entrepreneurial intentions: A meta-analytic review. Entrepreneurship theory and practice, 38(2), 217-254.
Bandura, A. (1997). Self-efficacy: The exercise of control. Macmillan.
Byabashaija, W., & Katono, I. (2011). The impact of college entrepreneurial education on entrepreneurial
attitudes and intention to start a business in Uganda. Journal of Developmental Entrepreneurship, 16(01),
127-144.
Bolton, D.L. & Lane, M.D. (2012). Individual entrepreneurial orientation: Development of a measurement i
nstrument. Education & Training 54 (2/3), pp.219-233.
Borchers, A., & Park, S. (2010). Understanding entrepreneurial mindset: A study of entrepreneurial self
efficacy, locus of control and intent to start a business. Journal of Engineering Entrepreneurship, 1(1), 51-62.
Cushing, E., Therriault, S., English, D. (2017). Developing a College- and Career-Ready Workforce An
Analysis of ESSA, Perkins, and WIOA
https://ccrscenter.org/sites/default/files/Career-ReadyWorkforce_Brief_Workbook.pdf.
Denning, S., (2016). https://www.forbes.com/sites/stevedenning/2016/05/27/why-us-entrepreneurship-is-dy-

ing/#788011be7d74
EPI, 2015. (http://www.epi.org/publication/the-need-to-address-noncognitive-skills-in-the-education-poli-

cy-agenda/)
Farrington, C. A., Roderick, M., Allensworth, E., Nagaoka, J., Keyes, T. S., Johnson, D. W., & Beechum, N. O.
(2012). Teaching Adolescents to Become Learners: The Role of Noncognitive Factors in Shaping School
Performance—A Critical Literature Review. Consortium on Chicago School Research. Chicago, IL.
Günzel-Jensen, F., Moberg, K., Mauer, R., & Neergaard, H. (2017). Self-Efficacy and the Entrepreneurial
Mindset Revisited. In Revisiting the Entrepreneurial Mind (pp. 319-335). Springer International Publishing.
Hayes, David, Richmond, William. (2017). Using an Online Assessment to Examine Entrepreneurship Student
Traits and to Measure and Improve the Impact of Entrepreneurship Education. Journal of Entrepreneurship
Education, 20 (1), 88-107.
Herman, J. L., Stevens, M. J., Bird, A., Mendenhall, M., & Oddou, G. (2010). The tolerance for ambiguity scale:
Towards a more refined measure for international management research. International Journal of Intercultural
Relations, 34(1), 58-65.
Hox, J. J., Bechger, T.M. An Introduction to Structural Equation Modeling. Family Science Review, 11, 354-373.
Hu, L.-t., & Bentler, P. M. (1998). Fit indices in covariance structure modeling: Sensitivity to
underparameterized model misspecification. Psychological Methods, 3(4), 424-453.
21 www.nfte.com
Hu, L.-t., & Bentler, P. M. (1999). Cutoff criteria for fit indexes in covariance structure analysis: Conventional
criteria versus new alternatives. Structural Equation Modeling, 6(1), 1-55.
Junior Achievement (2015). States of Entrepreneurship in America.

http://www.juniorachievement.org/documents/193855/1976054/States+of+Entrepreneurship+in+America.
pdf/8b15bcdd-1b5f-4d14-8f35-2fd130f4572b Kautz, T., Heckman, J. J., Diris, R., Ter Weel, B., & Borghans, L.
(2014). Fostering and measuring skills: Improving cognitive and non-cognitive skills to promote lifetime suc-
cess (No. w20749). National Bureau of Economic Research.
Keller, P. S., & El-Sheikh, M. (2011). Latent Change Score Modeling of Psychophysiological Data: An Empirical
Instantiation Using Electrodermal Responding. Psychophysiology, 48(11), 1578–1587.
http://doi.org/10.1111/j.1469-8986.2011.01225.x
Krueger Jr, N. F., Reilly, M. D., & Carsrud, A. L. (2000). Competing models of entrepreneurial intentions.
Journal of business venturing, 15(5-6), 411-432.
Krueger, Norris (2015). Entrepreneurial Education in Practice. Part 1 – The Entrepreneurial Mindset.
Entrepreneurship 360 Thematic Paper. Paris: OECD.
Krueger, N. (2009). Entrepreneurial intentions are dead: Long live entrepreneurial intentions. In Understanding
the entrepreneurial mind (pp. 51-72). Springer New York.
Li, CH. Behav Res (2016) 48: 936. https://doi.org/10.3758/s13428-015-0619-7
Lichtenstein and Monroe-White (2017). Entrepreneurial Mindset Assessment Reviews.

https://venturewell.org/wp-content/uploads/EMAR-v1-1.pdf.
McArdle, J. J., Hamagami, F. (2001). Linear dynamic analyses of incomplete longitudinal data.
In Collins, L., Sayer, A. (Eds.), Methods for the analysis of change (pp. 139-175). Washington, DC: American
Psychological Association
Millennial Branding (2012). http://millennialbranding.com/2012/millennial-branding-student-employ-

ment-gap-study/
Moberg, K. (2014). Assessing the impact of entrepreneurship education: from ABC to PhD.
Copenhagen Business School [PhD].
Moberg, Kåre. (2012). The Impact of Entrepreneurship Education and Project-Based Education on Students’
Personal Development and Entrepreneurial Intentions at the Lower Levels of the Educational System: Too
Much of Two Good Things?. SSRN Electronic Journal. 10.2139/ssrn.2147622.
Nakkula, M., Pineda, C., Dray, A., & Lutyens, M. (2003). Expanded explorations into the psychology
of entrepreneurship: Findings from the 2001-2002 study of NFTE in two Boston public high-schools.
Harvard University
Nunnally, J. C., & Bernstein, I. H. (1994). Psychometric theory (3rd ed.). New York, NY: McGraw-Hill, Inc.
Pihie, Z. A. L., & Akmaliah, Z. (2009). Entrepreneurship as a career choice: An analysis of entrepreneurial
self-efficacy and intention of university students. European Journal of Social Sciences, 9(2), 338-349.
Smith, K. (2017). http://digital.graphcompubs.com/article/Entrepreneurship+Empowers+

Students+With+Career+Readiness/2763554/400785/article.html.
Fornell, C., and Larcker, D. F. (1981). Evaluating Structural Equation Models with Unobservable Variables and
Measurement Error. Journal of Marketing Research (18:1), pp. 39-50.
Yu, C.Y. ( 2002). Evaluating cutoff criteria of model fit indices for latent variable models with binary and
continuous outcomes. Doctoral dissertation , University of California, Los Angeles
22 www.nfte.com
APPENDIX A
Model Fit, Reliability, Internal Consistency, Convergent and Discriminant Validity

Table 1 in the text and Table A-2 below provide a detailed description of our model fit indices as well as
estimates that detail reliability, internal consistency, and convergent and divergent validity.
Notable highlights from this analysis include the following:
• We were able to replicate findings from the testing phase, which showed that the EMI’s constructs
had high levels of reliability, internal consistency, and convergent validity. Our constructs exceeded
the recommended Cronbach’s alpha and composite reliability (CR) cut-off of 0.70 and average value
extracted (AVE) values of 0.5.45
• The only exception to this was our communication and collaboration domain in the pre-EMI, which has
the fewest number of items (three). However, even this construct improved to the acceptable threshold
of internal consistency in the test of the post-EMI assessment.
• We also found the six constructs collectively to be reliable and internally consistent, as assessed by
Crobach’s Alpha and the Fornell and Larcker recommended values. This is important given that our
later analysis examines overall EMI scores, instead of focusing on the six constructs individually.
• We were unable to establish discriminant validity across our constructs as the co-variance between
constructs was fairly high. To evaluate discriminant validity, we used Fornell and Larcker’s method
of comparing the square root of AVE values with the highest correlation between constructs. In
particular, there seemed to be a high correlation between the Creativity and Innovation and Critical
Thinking and Problem Solving constructs, as well as the Creativity and Innovation and Opportunity
Recognition constructs.
• Finally, the overall EMI model—which considers the six constructs jointly—was also a good statistical
fit, further supporting our decision to analyze the average entrepreneurial mindset score overall instead
of looking at the six constructs separately. We determined model fit using the fit indices laid out by Hu
and Bentler (1998, 1999); our comparative fit index (CFI) and Tucker Lewis Index (TFI) were above
.90, and our root mean square error of approximation (RMSEA) and standardized root mean square
residual (SRMR) were below 0.08. We also include chi-square estimates with the caveat that chi-square
is sensitive to large sample sizes.
23 www.nfte.com
Table A-1. Reliability, Internal Consistency, and Model Fit of Pre-EMI (Six Constructs)
24 www.nfte.com
Measurement Invariance
As noted in the text, we conducted measurement invariance analyses to understand how different subgroups
of students, in terms of ethnicity, gender, grade level and course requirement, approached the EMI questions.
These tests help us determine if there is any bias in the way we have framed the questions that would impact
the results.
In this analysis, we applied three different levels of testing: configural, weak, strong. Each level builds on the
previous one, adding additional constraints on model parameters to achieve strong forms of invariance at
each stage. Each stage is essentially testing increasingly restrictive hypotheses and further establishing the
measurement consistency of the different constructs.46
We noticed that some of our tests showed statistical differences for a few sub-groups as assessed by
chi-square (i.e., the chi-square statistic became significant from the configural to weak or from weak to the
strong invariance test). However, the changes in the CFI and RMSEA indices suggested that the model was
overall invariant between
sub-groups. Given that CFI and RMSEA have been recommended to assess model fit in large sample sizes,
we place greater emphasis on these model fit indices as well as the changes of these indices between
sub-groups.47
Tables A-2 – A-5 provide the model fit indices for all the sub-groups that we assessed for measurement
invariance in both the pre and the post-EMI. Generally, these tables show that students are interpreting the
EMI questions similarly, and that their responses are not affected by their ethnicity, gender, grade level, or
course status.
Table A-2. Measurement Invariance Testing for Ethnicity
25 www.nfte.com
Table A-3. Measurement Invariance Testing for Gender
Table A-4. Measurement Invariance Testing for Grade Level
Table A-5. Measurement Invariance Testing for Required v. Elective Course Status
26 www.nfte.com
Using Overall EMI Mean Scores
The bulk of our analyses in the previous section used factor scores, which was necessary to assess the EMI
instrument. Factor scores allowed us to examine correlations among items without any residual variance,
creating less statistical “noise”, as all analyses happened within the CFA model. However, for the rest of the
paper, we switch to actual observed mean scores of the EMI, primarily because we will use these scores for
final reporting purposes. Table A-6 on the next page displays the correlations between our pre-EMI observed
mean construct scores and the pre-EMI factor construct scores to establish a strong association between
observed mean scores and factor scores. This strong relationship between observed mean scores and factor
scores persists in our post-EMI.
Table A-6. Correlation between Pre-EMI Constructs Observed Means and Factor Scores

Observed Means
Factor Scores
27 www.nfte.com
We also assessed the relationship between individual constructs and an overall EMI mean score, comprised
of an average of the six constructs mean scores. Table A-7 on the next page shows that the correlation
between individual construct observed mean scores and an aggregate EMI mean score is also relatively
strong, especially in the post-EMI. The strong correlations between observed mean scores and factor scores,
as well as construct mean scores and an overall EMI mean score, enables us to conduct further analyses
using an overall EMI mean score.
Table A-7. Correlation between Individual EMI

Construct Means and an Overall EMI Mean Score
28 www.nfte.com
APPENDIX B
Table B-1. OLS Regression: Relationship between EM Change (Binary)

and ESE Change (Scaled)
29 www.nfte.com
Table B-2. Logistic Regression: Relationship between EM Change (Binary)
and ESE Change (Binary) – Odds Ratios
30 www.nfte.com
Table B-3. ANOVA Regression: Relationship between EM Change (Mean) and ESE Change
(Groups; 1 = negative change, 2 = no change, 3 = positive change)

31 www.nfte.com
Table B-4. Logistic Regression: Relationship between EM Change (Binary) and
Attitude Toward Entrepreneurship Being A Skill That Can Be Used in Any Career After NFTE
(Binary; 1 = Yes, 0 = No)
32 www.nfte.com
Attitude Toward Entrepreneurship Being A Skill That Can Be Used in Any Career
(Binary; 1 = Students Went from No to Yes, 0 = Students Went from Yes to No)
33 www.nfte.com
Attitude Toward Entrepreneurship Being A Skill That Can Be Used in Any Career
(Binary; 1 = Students Went from No to Yes, 0 = Students Went from No to No)
34 www.nfte.com
APPENDIX C
Table C-1. Logistic Regression: Relationship between EM Change (Binary)

and Desire to Want to Own a Business After NFTE (Binary; 1 = Yes, 0 = No)
35 www.nfte.com
and Desire to Want to Own a Business After NFTE
(Binary; 1 = Students Went from No to Yes, 0 = Students Went from Yes to No)

36 www.nfte.com
and Desire to Want to Own a Business After NFTE
(Binary; 1 = Students Went from No to Yes, 0 = Students Went from No to No)

37 www.nfte.com
APPENDIX D
Table D-1. ANOVA Regression: Relationship between EM Change (Mean) and Content Change
Ranks (Class-Level Groups; 1 = Highest Change Between Diagnostic and Summative,
4 = Lowest Change Between Diagnostic and Summative)
38 www.nfte.com
Table D-2. ANOVA Regression: Relationship between EM Change (Mean) and Content
Summative Ranks (Class-Level Groups; 1 = Highest mean score, 4 = Lowest Mean Score)

39 www.nfte.com
ENDNOTES
1
This is different from the emergence of the gig economy, which has led in recent years to an increase in the
number of freelancers. See Upwork’s 2017 report, “Freelancing in America.”
2
Denning, 2016.
3
Bacigalupo, Kampylis, Punie, & Brande, 2016.
4
See, for example, the World Economic Forum, 2016, “The Future of Jobs.”
5
Millennial Branding, 2012.
6
Bacigalupo et al., 2016.
7
Kautz, T. et al., 2014; Farrington, C. et al., 2012.
8
EPI, 2015. (http://www.epi.org/publication/the-need-to-address-noncognitive-skills-in-the-education-poli-
cy-agenda/)
9
Durlak, J. A., Weissberg, R. P., Dymnicki, A. B., Taylor, R. D. & Schellinger, K. B. (2011). The impact of
enhancing students’ social and emotional learning: A meta-analysis of school-based universal interventions.
Child Development, 82(1): 405–432.
10
Entrepreneurship programs are also growing outside of the traditional business departments and becoming
cross functional across universities. See https://www.nytimes.com/2015/12/29/technology/universities-race-to-
nurture-start-up-founders-of-the-future.html?_r=0
11
Junior Achievement, 2015.
12
Bacigalupo et al., 2016.
13
Comprehending the mindset of an entrepreneur is also of great interest to those who want to push for more
startups. A recent review of the literature suggests that a better understanding of the way entrepreneurs think
will provide more information about the antecedents to developing new ventures (Brannback and Carsrud
2017). Entrepreneurial mindset is one important factor in this equation, as it is a combination of cognitive
and noncognitive skills that, with the right context, will drive toward the development of new businesses.
14
Hayes and Richmond, 2017.
15
Borghans et al., 2008.
16
Lichtenstein and Monroe-White, 2017.
17
See VentureWell’s efforts to define entrepreneurial mindset:
https://venturewell.org/determining-entrepreneurial-mindset/.
18
Herman et al., 2010.
19
http://entremetric.com/
20
Bolton et al., 2012.
40 www.nfte.com
21
Emerging from the Danish Foundation for Entrepreneurship, the ASTEE project created an instrument
to assess the impact of entrepreneurship education on students by measuring entrepreneurial mindset,
knowledge and skills, as well as connectedness to education. The latter domain addresses the potential
impact of entrepreneurship education on a student’s academic mindset. The ASTEE project is further focused
on entrepreneurship education at multiple levels of schooling, from primary school to postsecondary.
22
NFTE is currently introducing an entrepreneurship pathway that takes students on a journey from awareness
and exposure to expertise and finally to application. For more on NFTE’s entrepreneurship Pathway,
see http://www.nfte.com/our-programs/.
23
Bae T.J., et al., 2014.
24
Pihie, Z. A. L., & Akmaliah, Z. (2009). Entrepreneurship as a career choice: An analysis of entrepreneurial
self-efficacy and intention of university students. European journal of social sciences, 9(2), 338-349.
25
Nakamoto and Rice, 2017.
26
Nakkula, et al., 2003.
27
Petway, K. T., Rikoon, S. H., Brenneman, M. W., Burrus, J., & Roberts, R. D. (2016). Development of the
Mission Skills Assessment and Evidence of Its Reliability and Internal Structure. ETS Research Report Series,
2016(1), 1-15.
28
We also explored using anchoring vignettes (AV), which are item types used to correct reference bias by
standardizing the varying reference points of test-takers when answering Likert scale questions. In the end,
we decided to not move forward with AV type items due to their wordy nature and heavy cognitive load that
made the assessment unwieldy.
29
For example, we found in one question for the flexibility and adaptability construct used the term “shifting
demands,” which the students did not understand. As a result, we changed the wording.
30
Situational Judgement Tests (SJTs) are a type of psychological assessment that present test-takers with
scenarios for which they have to rank different responses to each situation.
31
While we do not delve deeply into the initial testing phase of the EMI in this paper, the authors are happy
to provide interested readers with more information on the process and statistical techniques used by NFTE
and ETS.
32
For more on latent constructs, see https://johngarger.com/articles/methodology/latent-constructs-in-so-
cial-science-research.
33
Hox and Bechger, 1998.
34
These analyses use observed mean scores rather than factor scores; see Appendix A for more information.
35
Given our overall large sample size, our sub-group analyses contained at least 450 student responses with
individual cell sizes exceeding 170 observations
36
See, for example, Chen et al., 1998; McGee at al., 2009; Nabi et al. 2017.
37
Chen et al., 1998.
38
Mauer et al., 2017.
39
Bandura, 1977, p. 29, cited in Maurer et al, 2017.
40
Pihie, Z. A. L., & Akmaliah, Z., 2009.
41
Krueger, N. et al., 2000.
41 www.nfte.com
42
Given the number of students who experienced no ESE growth between the beginning and end of NFTE,
we also considered EMI mean change scores in context of the three ESE categories (negative growth, no
growth, positive growth). We found that EMI change scores were statistically significantly positive for students
with positive ESE growth and that the absolute EMI change score was also the highest for students with
positive ESE growth. Students with no ESE growth showed a negligible change in EMI between the
beginning and end of NFTE, whereas students with negative ESE growth showed a negative EMI change.
All of the above analyses included controls and state fixed effects.
43
Krueger, N. (2009). Entrepreneurial intentions are dead: Long live entrepreneurial intentions.
In Understanding the entrepreneurial mind (pp. 51-72). Springer New York.
44
In addition, positive growth in entrepreneurial mindset is a predictor for students who experience a change
in their desire to want to own a business. Students who grow in their mindset are more likely to change their
desire from wanting to own a business from “no” to “yes” compared to students who experienced negative
mindset change (see Appendix C).
45
Fornell and Larcker 1981, Nunnally and Bernstein 1994
46
For more on this approach, see
http://comm.eval.org/HigherLogic/System/DownloadDocumentFile.ashx?DocumentFileKey=63758fed-a490-43f
2-8862-2de0217a08b8.
47
Yu, 2002.
42 www.nfte.com
About NFTE
NFTE (Network for Teaching Entrepreneurship) is an internationally recognized
nonprofit that activates the entrepreneurial mindset in young people. Research
shows that the entrepreneurial mindset—skills including innovation, self-reliance,
comfort with risk, communication, and problem-solving—prepares young people
for lifelong success. Equipped with the mindset, as well as the business and academic
skills that NFTE teaches, NFTE students are ready to thrive in the innovation economy
no matter what path they choose. NFTE focuses its work on under-resourced
communities. Through its nationwide network of program offices, NFTE partnered
with 1,882 schools and community-based organizations across the U.S. last year,
reaching more than 52,000 students in 20 states. Partnerships in eight other countries
allowed NFTE to serve an additional 53,000 young people internationally.
Learn more at www.nfte.com.
About NFTE’s EMI Research

For more information about NFTE’s Entrepreneurial Mindset Index, contact
Sophia Rodriguez, Director of Research & Analytics, at sophia.rodriguez@nfte.com.

Measuring Entrepreneurial Mindset in Youth

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Measuring Entrepreneurial Mindset in Youth

Uploaded by

Copyright:

Available Formats

MEASURING ENTREPRENEURIAL

Thomas Gold and Sophia Rodriguez

PART 1—WHY ENTREPRENEURIAL MINDSET MATTERS........................................4

PART 2—THE ENTREPRENEURIAL MINDSET INDEX (EMI).....................................8

PART 3—EARLY EMI RESULTS:

Why Entrepreneurial Mindset Matters

Developing the EMI

Early EMI Results: Exploring Changes in Entrepreneurial Mindset

Entrepreneurial Mindset Index Domains

Building Skills that Employers Need

Teaching Key Noncognitive Skills that Help Students Succeed

The Next Challenge: Effectively Measuring Entrepreneurial Mindset in Youth

Focus of this Paper

Entrepreneurial Entrepreneurial Entrepreneurial Entrepreneurial

Deciding What to Measure

Developing and Piloting Items

Model Fit, Reliability, Internal Consistency, Convergent and Discriminant Validity

About the Sample

Figure 3. Entrepreneurial Mindset Index Results, 2016-2017

How Do EMI Scores Relate to Other Relevant Entrepreneurial Attitudes

• Are twice as likely to have positive entrepreneurial self-efficacy (ESE) growth

EMI and Entrepreneurial Self-Efficacy

EMI and Entrepreneurship Attitudes and Intentions

Bandura, A. (1997). Self-efficacy: The exercise of control. Macmillan.

Denning, S., (2016). https://www.forbes.com/sites/stevedenning/2016/05/27/why-us-entrepreneurship-is-dy-

EPI, 2015. (http://www.epi.org/publication/the-need-to-address-noncognitive-skills-in-the-education-poli-

Junior Achievement (2015). States of Entrepreneurship in America.

Li, CH. Behav Res (2016) 48: 936. https://doi.org/10.3758/s13428-015-0619-7

Lichtenstein and Monroe-White (2017). Entrepreneurial Mindset Assessment Reviews.

Millennial Branding (2012). http://millennialbranding.com/2012/millennial-branding-student-employ-

Smith, K. (2017). http://digital.graphcompubs.com/article/Entrepreneurship+Empowers+

Model Fit, Reliability, Internal Consistency, Convergent and Discriminant Validity

Notable highlights from this analysis include the following:

Table A-2. Measurement Invariance Testing for Ethnicity

Table A-4. Measurement Invariance Testing for Grade Level

Table A-7. Correlation between Individual EMI

Table B-1. OLS Regression: Relationship between EM Change (Binary)

Table C-1. Logistic Regression: Relationship between EM Change (Binary)

About NFTE’s EMI Research

You might also like